Monday, 9 February 2015

Dedomestication series pt. IV: Implications for "breeding-back"

The aurochs was, per definitionem, the wild type of Bos primigenius. If “breeding back” aims to approach the aurochs as close as possible, the result has, ultimately, to be a wild animal. Artificial breeding with domestic breeds can only result in an animal that is itself domestic, no matter how aurochs-like it looks, and it will still display a number of domestic traits – be it optic, behavioural or whatever.

For part I, for part II, for part III, for part IV.

Nature knows better

Our knowledge of the developmental cascades of many domestic traits is probably not deep enough to select against them and reconstitute the original, wild-type state. See for example the fact that selective breeding in farm foxes for earlier maturity and more offspring directly was not fruitful, but selection for tameness alone brought exactly that result. How to select on the genes coding for endocrinologic cascades that regulate the amount of hormones that control how bulky or muscular the body is, or those that regulate the timing of developmental mechanisms that result in an either elongate or paedomorphic skull? I am sure that just selection for long snouts or an athletic body directly in a population where these features are present only to a mediocre extent (f.e. uncrossed Heck cattle) will not be fully successful. The same goes for achieving a well-developed sexual dimorphism both in size and colour. Just selecting for large bulls or small cows certainly would not do it, breeding only with bulls that mature late would slow down the whole breeding process which is slow enough already and it even is not sure whether this is the right way or not. If just always taking out all black cows is the effective way to achieve a fixated, well-pronounced sexual colour dimorphism is dubious as well, because always even in Heck cattle herds with a good sexual dimorphism half of the cows are coloured like bulls.

Artificial selection might even be counteracting itself. I already expressed my thought that the muscular, athletic body of Lidia might be linked to its temperament. For example, the True Nature Foundation plans to do a project trying to collect suitably aurochs-like Lidia and breed them for a greater resemblance of their ancestor (which is awesome, I have been dreaming of such a project for a long time). But the selection program will also include selection against aggressiveness to make them easier to handle. That is fully understandable, grazing projects do the same, nobody wants dangerous cattle that are impossible to handle. But actually this is the same breeding for tameness and docility as in the farm fox experiment, so maybe it will result in a further domestication of fighting cattle and therefore reduce the athletic, muscular body and cause more paedomorphic features? This is just an idea, future might show if there is truth in my guess. The same goes for the pleiotropic connections between traits, such as those that cause white spots. I think that most of those connections are hardly accessible to us and that hinders us from effectively selecting for the desired phenotypic traits. For example, domestication in bovines almost always results in the horns getting “pulled” outwards or up/downwards. This is apparent in a lot of cattle and sheep breeds. Many breeds used in breeding-back have such horns and it proved to be difficult to select for a fully aurochs-like horn shape, but in OVP it seemingly developed by itself in a few individuals. The next problem is that we do not know how large exactly the impact of phenotypic plasticity is (f.e. on muscularity of the body, horn and body size, skull shape etc.; in the case of behaviour it is clear that environment is a crucial factor). To evaluate that, extensive comparative studies would be necessary which would take several years.

Probably all cattle would show herding and defensive behaviour in the wilderness. Free-ranging cattle do and also farm cows defend their calves aggressively. It is open whether the reproductive circle of cattle adapts to the seasons due to environmental impulses or natural selection, but I think the latter is the case because cattle in grazing projects are not fully adapted to the seasons in this respect. Perhaps natural selection could be mimicked by taking out cows and their calves that calved during winter or fall, but I am not sure if this would work.

All in all I would simply say that “nature knows better” in any case. This does not only apply to ecologic and immunological capacities, which are not visible and already present in primitive landraces to a certain(!) extent. The fact that certain wild traits appeared in free-ranging populations whose founding individuals certainly did not have them (f.e. the hump and body conformation in OVP, S.v.F. and Amsterdam cattle, or the long snout in some cows at OVP) and probably would have been hardly achieved by selective breeding implicates to me that natural selection “re-develops” aurochs-traits in the way outlined in the previous posts and better than artificial selection ever could. Artificial selection probably would be more of a mimic of adaptive wild type (=aurochs) traits, while natural selection produces truly adaptive traits. Natural selection knows better how to eradicate domestic traits such as paedomorphy or a reduced sexual dimorphism because we have no exact clue of the connections and interactions of development, pleiotropy and environment in morphology, behaviour and ecologic and immunologic capacities.

How to do it

Nevertheless, I think that simply tossing some bunch of any cattle into nature will not result in the dedomesticated near-aurochs that we want to see. The existing free-ranging cattle populations demonstrate that. If the founding individuals do not have horns that are either already aurochs-like in certain aspects or at least very diverse, it would take either very long for aurochs-like horns to evolve or they would end up in a more or less different shape. Aurochs colour (E+) won’t evolve if not present. Colour probably only has a weak influence on evolutionary fitness in cattle and probably coincidence will play a large role in when a certain colour variant becomes fixated in the population. So releasing just any cattle of any colours will not result in an authentic, aurochs-like colour – just as an example.  

As outlined in pt. III, we cannot simply expect that “nature” always changes the cattle in the way the aurochs was. Evolution is not static but dynamic and animals always adapt to the current circumstances. If we want a wild, near-aurochs, we have to simulate the evolutional adaptive environment the original lived in. The European aurochs evolved under the predative pressure not only of wolves (as juveniles also lynxes, foxes and bears), but also big cats (not only until the end of the Pleistocene; lions and leopards still lived in parts of southern Europe into the antiquary). They had way more space to live, graze and migrate and had to compete with a whole range of other herbivores. For that, a large reserve would be necessary to avoid the island effect and introduce competing herbivores (the most important competitors probably are deer and horses, as realised in the OVP) and carnivores. The introduction of big cats is a topic that I do not want to open here but let’s assume it won’t happen and there will be only fox, lynx and wolves (bears are problematic as well). No medical care or supplementary feeding will cause legal problems which should be solved like in the OVP and what the Tauros Project is trying to achieve, i.e. to classify the cattle as res nullius and legally wild animals. The cattle would have to be totally reproductively isolated, otherwise the dedomestication process would slow down.

And that’s how I would do it:

At first I would try to produce an aurochs effigy that is as authentic as possible by selective breeding (“breeding-back”). All or most desired traits that can be achieved should be present and as be stable as possible. Then I would introduce them together with individuals of primitive cattle like good Sayaguesa, Lidia, perhaps Maronesa and good Chianina, Castellana Axarquica, good Boskarin, Betizu for their feral history and Yakut cattle for their great adaptions to cold and their genetic distance to European breeds. The result would be a genetically diverse population that has all aurochs features and all founding breeds well-adapted to harsh climate, sparse vegetation and resistant to diseases. But I would not include too many individuals of breeds with strong dilution factors and/or short horns. The heterogeneity in the first generations would be high, and it would certainly be very interesting to see how the frequency of the single features is going to evolve. The traits should be categorized and their frequency be evaluated all three years or so, to document the evolutional shift in the population. Selective culling should only be used as a tool when it is apparent that natural selection has left a strong mark in the population already. Culling should focus primarily on fur colour, and only be carried out if the population is in a good state. For example, the Heck cattle at OVP are in a crisis at the moment and the last thing that I would do now is selective culling. I think an ideal population size would be 500 individuals at least. In any case more than 100.

But let’s be realistic, an area large enough to sustain a viable population of all the three herbivores plus wolves and perhaps also lynx that also can remain totally untouched by human management is not easy to achieve for conservation. But also without predators, most of the selective pressures described in part II would still be there, and if you have read carefully you probably noticed that predators are likely not among the most important factors acting upon such a free-ranging population.

The concept I described is basically what the Polish naturalist Feliks Pawel Jarocki suggested as early as 1835 (without the selective breeding part), only eight years after the aurochs was formerly described. He proposed that the release of cattle into wilderness so that they would live under the same circumstances as the aurochs did, would result in a “revival” of the original form. Note that I am not claiming that the original European aurochs can be revived the way described here.

Therefore what I am suggesting is a mix of breeding-back and dedomestication. My opinion is that dedomestication inevitably has to be the end phase of creating a near-aurochs. “Natural selection with a kick-start”, as the Tauros Project would call it.

Feral, wild and dedomesticated

At which point should be call such a dedomesticated aurochs-like population “wild”? Actually I have been dealing with terms like “dedomesticated”, “feral” and “wild” without defining them appropriately all the time. What I do now is what I should have done right at the beginning of the dedomestication series, I apologize. Parts of my definitions are inspired by those given by the user “Roberta” in the Carnivora Forum thread on the aurochs.

While genetics use the “wt” terminology mainly for alleles or any phenotypic features, zoology considers all aspects of the entire species, also including its history and interaction with the environment. Both feral and wild populations have in common that they are not enclosed (on small scale) and not dependent on active human help (this does not apply if the species is a commensal, where its ecological niche is living from and around human civilization). A wild species evolved/-s with and within its ecosystem and therefore does not change it rapid and dramatically. A feral species, on the other hand, is sometimes invasive and sometimes not.

Another distinction between feral and domestic might be the amount of genetic structure that was either influenced by evolution or by man. Considering that there are actually a lot of populations of wild animals are managed to some extent and are physically limited, this aspect might be of much greater importance. A wild type trait is defined as a trait that is shaped by evolution, not by man, and occurs in nature. As you have probably noticed, this applies to both a pre-domestic and a (I think hereby I introduce a new term) post-domestic phase. Classic zoological terminology only recognized pre-domestic wild animals as “wild-types” and therefore a wild type is usually also understood as the wild forerunner of the domestic type, and it never was considered that there can be a post-domestic wild-type as well. If you will, the Dingo or the European mufflon are post-domestic wild animals to me, or at least wild animals. They have been living in the wild for long enough, are adapted to their environment (with the exception of those mufflon populations that have been introduced where this species simply is not native, f.e. wet Central European lowland forests) and are an integral part of the ecosystem without causing changes that we call “damage”. But as you see, the line between feral and wild becomes arbitrary at this point. I would call Chillingham cattle and the Heck cattle at OVP not even truly feral, because both live on a rather small, enclosed area and furthermore, the Heck cattle have not been living there for all too long, and the genetic structure of Chillingham cattle is highly influenced by man (see previous part). In this case, I prefer to call the state they are living in “free-ranging”. Amsterdam Island cattle and Betizu on the other hand are/were feral in my understanding of the word.

No question that a post-domestic wild animal has to be a dedomesticated animal. Dedomestication is, as the word implies, the “reversal” of domestication. However, I prefer to call it the loss of domestic traits, as “reversal” implies the full re-emergence of the original state, which is not the case. Those domestic traits are, to put it in a nutshell:

- Paedomorphic features both in behaviour and morphology.

- reduced sexual dimorphism and usually also brain volume

- novel morphological traits (colour variants, horn shapes/sizes, change of body size, appendages etc.) due to pleiotropy, developmental cascades and relaxed selection; sometimes exaggerated due to artificial selection

- reduction of traits crucial for reproductive success – behavioural, immunological, sensory, ecological or morphological – due to relaxed selection

- loss or reduction of traits linked to sexual selection due to artificial selection

- Heterogeneity due to genetic drift and artificial selection

I would not say that a post-domestic or secondary wild type has to have lost all of those domestic traits; only those that lower the reproductive success. If natural selection or genetic drift allows or even fixates novel traits like some colour or horn variants or structures like a large dewlap, I think it is OK. In my opinion, they don’t even have to be completely homogeneous in terms of horn shapes and colour. First of all, the aurochs was quite diverse regarding the pronunciation of the “primigenius spiral”, size and orientation relative to the snout, and the snout length itself varied slightly as well. Although most wild species are highly uniform, there are some displaying more than one colour variants. Northern wolves for example are more or less diverse in that respect, partly also thanks to domestic dog introgression.

How to scientifically categorize the such a type of cattle? Regardless of if you regard domestic animals taxonomically relevant or not, these cattle would not be domestic, so why not classifying them? If so, I would simply tag them as Bos primigenius taurus because they will be neither genetically nor phenotypically 100% identical to any of the three aurochs subspecies, and the epitheton “taurus” has already been given for the domestic cattle clade. However, a prerequisite for a taxonomic status is that these post-domestic cattle form one reproductive (meta)population.

All in all, the line between feral and wild is fluid and the distinction is subjective. I doubt that we will live long enough to see a fully dedomesticated near-aurochs, but at least there are good chances to see an aurochs-like cattle population in a process of dedomestication which shows clear signs of evolutionary changes. The challenge is to find a suitable area of sufficient size and overcome legal and public issues.


  1. Here's what I think could be interesting. "Breeding-Back" and "de-domestication" do not need to be performed sequentially, but may benefit from being done in parallel. For example, you could start two completely separate cattle populations using a diverse selection of primitive breeds for each. Every few years you would then rank each adult; the back-breeding herd is ranked using a predetermined aurochs effigy scoring system, while the de-domestic herd is ranked using some sort of health scoring system. You would then cull the lowest 1/3 of each herd each time you perform a ranking, using their separate ranking systems. The top 1/3 of each herd according to their unique ranking systems would be left alone. The middle 1/3 would be swapped between the two herds. Over time, the de-domestic herd should grow to look more like an aurochs faster than nature would otherwise dictate, while the back-breeding herd would come to act more like an aurochs than selective breeding would otherwise dictate. Furthermore, it would allow for novel combinations of genetics that may better unlock some of the developmental cascades.

  2. " lions and leopards still lived in parts of southern Europe during the antiquary"

    It is well established that last European lions were killed off in the Balkans around 0AD, but I haven't seen any proof of post-pleistocene European leopard. Is there really proof for that?

    Though I wouldn't be at all surprised if leopard outlived lion in Europe. Smaller, less aggressive towards humans and their livestock, more adaptable and so on. It should have adapted better in the diminishing prey during early holocene.